Upload
777894292
View
213
Download
0
Embed Size (px)
Citation preview
7/27/2019 IJETAE_1112_64
http://slidepdf.com/reader/full/ijetae111264 1/7
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, Volume 2, Issue 11, November 2012)
408
Calculation of the Factor of Stress Concentration in theJunction’s Welded Tube Subjected to Combined Loadings
A. Fouathia1, A. Mekroud
2, A. Benmeddour
3, K. Bellagh
4
1,2,3,4 Mechanics Laboratory, Faculty of Engineering, University Mentouri-constantine Chaab Ersas Campus, 25000
Constantine Algeria
Abstract — Fatigue failure caused by stress
concentrations in welded tubular joints is observed in the
offshore platforms subjected to cyclic loading in corrosive
marine environments. Therefore, it is necessary to
accurately assess the intensity of stress concentration to
properly address the problem of fatigue damage, and result
in reliable tubular junctions. Stress concentrations in theoffshore structures usually occur at the intersections of
tubular joints. In some junctions, the stress concentration
may induce a maximum stress at the intersection 30 times
the nominal stress, and increase the risk of fatigue failure in
tubular joints. This work aims to study the stress
distribution points and "hot" in the tubular welded joints, subjected to combined loading of traction and rotating
bending (bending in the plane, out of plane bending and
traction)simulating better loading real.
Keywords — stress concentration, offshore structures,
finite element method; tubular welded joints.
I. I NTRODUCTION
Welded tubular assemblies are widely used in steel
construction in the modern bridges, towers and offshore
structures such as oil rigs Offshore Jacket style for the
exploitation of hydrocarbon reserves in the marine
environment.They are classified according to their form
in T. Y. X, K, DT, DY, and DK [1].
The welded the frames with the amounts and diagonalforms a node characterized by a variable stiffness, non-
uniform distribution of stresses and a complex three-
dimensional behavior, while this may be unfavorable if
the assembly is poorly designed or poorly made and can
lead to fatigue failure resulting from cyclic loading to
which the structure is subjected[2]. Offshore platforms
are subjected during their service life (20-30 years) to
various environ mental actions; waves, currents, wind
and storms are particularly severe suffering and the
combined action of several stresses including traction
[3]. Bending in the plane and bending out of plane, the
stresses at the junctions of tubes create hot spots or areas
of high stress concentrations [4]. In some junctions, the
stress concentration may induce a maximum stress at the
intersection 30 times the nominal stress inevitably
leading to a fatigue damage of these structures [5].
Besides the security problem they pose, the fatigue
failure cause significant operating losses for the user andcostly to repair and redesign high for the manufacturer
[6].
It is therefore necessary to assess accurately the
intensity of stress concentration to properly address the
problem of fatigue damage, and result in reliable tubular
junction, and it will pay particular attention to the design
and the achievement of welded assemblies.
So far, research [7] in the field of fatigue weldedtubular nodes was conducted mainly conducted by the
offshore oil industry[8].
This study discusses the phenomena of fatigue in the
nodes of T tube assemblies welded metal structures
subject to marine stress due to random natural elements
(waves, wind, current, ...)
Its purpose is to study the stress distribution and
location of points "hot" (hot-spot stresses) at critical points in terms of fatigue in welded tubular joints
subjected to static loading along the traction, bending in
the plane, bending out of plane and the combination:
tension / bending in the plane and traction / bending out
of plane loading simulating more real it is to find anumerical simulation tool to study and predict the
behavior of welded tubular structures, a simple and
accurate modeling of the structure using a computer code
based on the finite element method that will calculate the
stress concentration near the weld. A mixed formulation
will ensure a balance of efforts at the junctions locally
and globally, on the whole structure.
II. R ESULTS A ND DISCUSSION
1 Stress distribution
The constraints considered in this study will be
calculated at the centroid of each finite element, so as tohave an average value of the stress tensor finite element
considered. These are the Von Mises stresses.
The results of the simulation of the load of the tubular
structure welded in tension we can take the existence of
two zones in the spacer:
• area 1 where the stresses decrease abruptly, is located
above the weld and in the immediate vicinity of the latter,
• area 2 in which the stresses gradually decrease, while
having a small change from place to reach values nearly
constant at the upper end of the brace, figure 1.
σN2 is the nominal stress measured in zone 2.
σN1, σN2: nominal stresses sampled at the edge of the
“hot” area located on the extension, above the two hotspots related to two points of crown point at a height ofabout (D + d) / 2 from the point of saddle.
7/27/2019 IJETAE_1112_64
http://slidepdf.com/reader/full/ijetae111264 2/7
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, Volume 2, Issue 11, November 2012)
409
Σmax is chosen as the highest values of σN1, σN2.
Either: σmax = sup (σN1, σN2) (1)
Figure1: Areas of stress distribution
In figure 2, which shows the evolution of the Von
Mises stress along the periphery of the brace, there is the
presence of the two areas already explained:
Figure 2: Distribution of stress on the brace.
By applying a traction force on the brace, the structure
undergoes a deformation whose general form is shown inFigure3, we see that the deformation is largely located in
the middle of the chord just below the brace; the latter
also undergoes a strain but lower proportion to the chord.
The deformation of the latter is mainly longitudinal.
Figure 3: Field of stresses in the structure with distorted.
2.1 Constraints in the junction
In the case of traction stress concentration around the
junction is divided symmetrically into two points and
this, it is largely visible in the vicinity of two points of
the crown favors the appearance of the hot spot in the
two points. In Figure 4 we see this stress concentration.
The maximum point is 32.4 Mpa High Chairs and the
minimum value is 0.11 Mpa.
Figure 4: Concentration of stresses in the Weld.
For reasons of symmetry, we consider only half of our
welded structure, so the study will focus on the portion
between the saddle point (Φ=0°) and crown point
(Φ=90°).
The figures, 5a and 5b show the distribution of stress
intensity factor kt in the structure. It is clear that the
concentration increases gradually from the point of
saddle until it reaches the maximum value at the crown
point (kt = 6.4). In Figure 5b best seen changes in the
stress concentration factor based on the positioning angle
Φi between two points saddle and crown, this explains
the increase in concentration.
7/27/2019 IJETAE_1112_64
http://slidepdf.com/reader/full/ijetae111264 3/7
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, Volume 2, Issue 11, November 2012)
410
Figure 5 a: stress intensity factor in traction
Figure 5b: Change kt according Φi
2.2 Out of plane Bending
In this case, the structure will be stressed in bending
out of plane this solicitation will result in a transverse
deformation which will be located primarily on the
frame, with a party that will undergo a pull and the othera compression figure 6.
Figure 6: stress concentration in bending out of plane
2.2.1. Constraints at the junction
In this case load, we can distinguish a similarity in the
distribution of stress concentration with the traction that
is to say that the concentration is on two points of crown,
see figure 7. The concentration changes in the same way
as in the case of traction, except that the values are lower
compared to those of traction, again for reasons ofsymmetry we consider only the portion between the
saddle point and crownpoint.
The minimum value is always at the saddle point at the
angle Φ=0°, while the maximum is at a crown point
(Φ=90°), and is in the range of 296.7Mpa.
Note that in the case of OPB, the value of the stress of
the hot spot is larger than that of the traction.
Figure7: Distribution of stress in the welded joint.
The distribution of stress intensity factor kt throughout
the tubular structure is representing in figure 8a. The
curve in figure 8b representing the evolution of the stress
intensity factor kt based on the angle of position Φi, takes
the same shape as the curve of traction. Concentration
increases gradually from the saddle point until it reaches
the maximum value at the crown point about 7.8 Mpa.
Figure 8a: stress intensity factor in traction
7/27/2019 IJETAE_1112_64
http://slidepdf.com/reader/full/ijetae111264 4/7
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, Volume 2, Issue 11, November 2012)
411
Figure 8b: Curve of variation of kt based Φi
2.3. Bending in the plane
We apply in this case a bending in the plane of the
frame creating a deformation made by compression on
one side and pulling the other. Note that most of thedeformation is located on the chord near the junction
between the two tubes, the spacer undergoes
deformation, but also to a lesser extent, figure 9.
Figure 9: Distribution of stresses in the structure
2.3.1. Constraints at the junction
From Figure 10, which shows the distribution of
stresses in the structure (the case of IPD) unlike the two
previous cases of loading, the stress concentration is notlocated very near the crown but is diverted to the saddle
point. It is in the range of 141.6Mpa.
In both previous cases we distinguish the presence of
two hot spots; in the case of the IPD we also notice the
same thing.
The symmetry in the distribution of the concentration
still exists.
Figure 10: Concentration of bending stresses in the plane.
The concentration in this case, changing gradually
from the saddle point to a maximum value (kt = 5.06) at
an angle Φ=90° and then begins to decrease until crown
point. We note here that the hot spot is not located at the
crown point as the other two cases, but in the saddle
point figure 11.
Figure 11: Factor intensity of traction stresses.
In the same way as in the other two loads we draw the
curve representing the evolution of the stress intensity
factor based on, but this time the shape of the curve is
completely different from the other two. But we can see
the appearance of symmetry of this curve.
7/27/2019 IJETAE_1112_64
http://slidepdf.com/reader/full/ijetae111264 5/7
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, Volume 2, Issue 11, November 2012)
412
Figure 12: Curve of variation of Kt based Φi Solicitation of the
structure IPD
2.4. Comparison of results
In Table 1 we compared the results of our calculationsand those obtained by parametric formulas and numerical
calculations of different researchers [9], a structure that
represents the same dimensions.
Load
cases
Kuang
& al.Gibstein
Hel.
& alF.E.M
our
calculat-ion
TRAC
-TION 9.6 11.2 6.7 8.6 6.48
IPD 3.3 3.4 2.8 3.2 5.06
OPD 8.6 11.3 6.1 9.1 7.81
Table 1: Comparison of results
We find that the results of our calculations are closer
to those obtained in particular by Hel. and al, in general,
by other researchers, for the same loads.
III. COMBINED LOADING
3.1. Combination pull-OPD-IPD
Applying a combined loading of traction and bending
out of plane and bending in plane with the same load of 4
Mpa. Note that the stress concentration at the junction
between the chord and the brace is located in the vicinity
in two points of crown. Given the combination of loads,
the chord is extremely distorted.
Figure 13: Concentration of stresses in traction-IPD-OPD.
Figure 14: distribution of stress in the welded joint.
This stress concentration is not symmetric. It startswith a maximum value 6.3 to crown point, Φ=0° angle
and begins to increase until it reaches the value of 10.3 at
Φ=72° angle, then the constraints will gradually decrease
until it reaches the value of kt = 3.6 to angle Φ = 143 °,
from that point until it increases the value of 6.3 kt in the
second crown point at the angle Φ=180°.The graph representing the evolution of the stress
concentration factor kt based on the positioning angle is
shown in Figure 15.
Figure 15: Factor intensity of traction
7/27/2019 IJETAE_1112_64
http://slidepdf.com/reader/full/ijetae111264 6/7
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, Volume 2, Issue 11, November 2012)
413
Figure 16: Curve of variation of Kt based Φi.
IV. DISCUSSED A ND R ECOMMENDATIONS
4.1. In case traction
Constraints prevailing at the junction chord-brace in a
tubular T as an axial stress produced by bending, due to
the overall deformation of the chord. Any changes to
these constraints will have a significant influence on the
values of the Kt and reducing the deformation of thechord.
The best way to reduce the deformation of the chord is
to add a stiffener where the chord is deformed. Research
[10] [11] has proven that this method can reduce themaximum hot spot up to 40%. Following his research,
Baker gardening came to reduce the maximum hot spotof 5.25 to 3.15, using this principle.
4.1.1. In case of bending in the plane
In the case of bending in the plane, the stresses at the
junction spacer ribs are produced by local bending of the
chord.
By acting the same way as in the case of traction, that is
to say by adding a stiffener in the chord, you can get a
slight reduction in the maximum value of kt, however,
this value is low given the meaning deformation of the
chord.
4.1.2. In case of bending out of planeBending out of plane is similar to that of traction when
the overall deformation is the main parameter generating
constraints at the junction spacer chord.
As in the case of traction, any reduction in the overall
deformation results in reduced values of the Kt.
With the same principle of traction as stiffening, themaximum values of kt can be reduced to 30%.
The calculation of stress concentration factors can
provide some information on the distribution of stresses
in the structure studied in the type of solicitations.
Given the complexity due to many parameters and in
particular the geometry of the intersection curve, there is
no straightforward method to take into account the stressdistribution near the weld, only the method of FEA can
make any part of the answer.
Some research work, namely those conducted by the
CTICM [12] (Industrial Technical Centre of MetalConstruction, France), have established the variation of
stress concentration factor according to each geometric
parameter, as a function exponential type.
The work carried out on cylindrical notched loaded in
tension, have to take into account the exponentialincrease of the stress distribution.
The fact remains that this possibility is being ruled out
of tubular structures, except in tension [13].
Therefore, the expression of stress concentration factor is
given by the ratio of the maximum stress σmax to the
hottest point on the stress nominal σnom near the weld, both on the side of the spacer as that of the chord. We
consider for each finite element of the structure, and for
each stress, the Von Mises stresses, in order to obtain the
stress concentration factors respectively. Constraints are
taken into account the constraints of so-called "skin" for
thin shell elements [12].But we can say to the existence of a zone of symmetry
between the saddle point and the crown point; in this area
the factors of stress concentration are identical, and are
found to increase with as we advance the saddle point to
crown point. The hot spot changes position depending on
the geometry of the node and the nature of the
solicitation.
V. CONCLUSION
Use of COMSOL Multiphysics code calculation based
on the finite element method to study and predict the
behavior of the tubular structure welded T-shaped a
simple and accurate modeling of this software. To
calculate the stress concentration (the factors of stressconcentration) in the vicinity of the weld and also to
locate hot spots or areas of high stress concentrations.
Given the complexity and nodes tubular geometric
discontinuity, this method seems best suited,
This study has shown that hot spots are generally located
in the crown points: traction/bending out of the plane and
combined traction / bending out of plane with values ofthe factors of stress concentration which can reach
respectively: ( 8.99, 10.36, 13.98) and are satisfactory
compared to those found by other researchers (9.62,
10.45, 18.0) [13][14], except in the vicinity: in plane bending and the combination: traction/bending in the
plane with values respectively as : (4.94 to 58.5°, 11.01
to 72°) and that approach also values obtained by the
searches: (3.11 at 45°, 9.8 to 67.5°), based on these
values shows that the maximum stresses in the hotspots
also vary depending on the load. The computer code can be used to achieve consistent results and consistent with
the literature.
7/27/2019 IJETAE_1112_64
http://slidepdf.com/reader/full/ijetae111264 7/7
International Journal of Emerging Technology and Advanced Engineering
Website: www.ijetae.com (ISSN 2250-2459, Volume 2, Issue 11, November 2012)
414
REFERENCES
[1 ] F.Ghanameh, D.Thevenet, A.Zeghloul "Concentration de
contraintes dans les jonctions tubulaire de type T, Y et K" 17eme
congré de mecanique-Troys septembre 2005-
[2 ] Fathi GHANAMEH : "Fiabilite des jonctions tubulaires soudees
des plateformes offshore" (2002-2005)
[3 ] S. A. KARAMANOS, A. ROMEIJN, J. WARDENER, "Stress
Concentrations in Tubular Gap K-Joint: Mechanics and Fatigue
Design", Engng, struct. vol.22 , 4-14, 2000.
[4 ] VISSER W, "On the structural design of tubular joints", Offshore
Technology Conference, OTC 2117, Houston, Texas (1974).
[5 ] American Welding Society, "Structural Welding Code-Steel",
14th edn, ANSI/AWS, Miami (1994).
[6 ] American Petroleum Institute, "Recommended Practice for planning, Designing and constructing Fixed Offshore Platforms",
1st edn, API RP2a-LRFD, Washignton DC (1993).[7 ] Lloyd's register of shipping. "Stress Concentration Factors for
Simple Tubular Joints", HSE BOOKS, OTH 354 (1997).
[8 ] Shao Yong-Bo "Geometrical effect on the stress distribution alongweld toe for tubular T- and K-joints under axial loading" Science
direct Journal of Constructional Steel Research 63 (2007)
1351 – 1360
[9 ] A. N’Diaye, S. Hariri, G. Pluvinage and Z. Azari . “Stress
concentration factor analysis for welded, notched tubular T-jointsunder combined axial, bending and dynamic loading”
International Journal of Fatigue, Volume 31, Issue 2, February2009, Pages 367-374
[10 ] M.M.K. Lee, D. Bowness. “Estimation of stress intensity factorsolutions for weld toe cracks in offshore tubular joint ”.
International Journal of Fatigue 24 (2002) 861 – 875
[11 ] CEE 504: Finite Element Methods in Structural Mechanics Winter2009 Using COMSOL Multiphysics to solve structural mechanics
problems.
[12 ] Rapport CNEXO-CTICM, "Calcul Statique des Assemblages
Tubulaires, Concentration des Contraintes dans les Assemblages
Tubulaires".
[13 ] E. CHANG and W.D. DOVER, "Stress Concentration Factor
Parametric Equations for Tubular X and DT Joints", Int. J.
Fatigue, vol. 18,363-387, 1996[14 ] A.N’DIAYE, "Concentration des contraintes dans des structures
tubulaires" Thèse de doctorat (Mars 2001) Universite de Metz.